T-Line Trading: The 8-Day EMA Rule, How It Is Read and How It Did

The T-Line is the 8-day exponential moving average of the closing price, the 8 EMA, used as a trend line for swing trades. Hit and Run Candlesticks, the site run by Rick Saddler and others where the name comes from, says the term was coined “back when I was working as a moderator in a trading room in 2004”, and adds: “Of course, I am not the first person to use the 8 EMA.” The T-Line is a name and a trading plan built on a standard average, not a new indicator.

What is the T-Line indicator?

An exponential moving average gives more weight to recent closes than a simple average does. With 8 periods, the latest close weighs 2 / (8 + 1), about 22% of the value, so the line hugs price closely and turns within a few bars. On a daily chart, the T-Line is a short-term trend line: a close above it is a close above the recent weighted average of prices, a close below it the opposite.

The T-Line is not a 10-period line: the source of the name defines it as the 8 EMA, and a 10-period average is a different line.

How to trade with the T-Line

The Hit and Run Candlesticks page gives one basic rule: “if you are in a long position, you want the price to close above the T-Line to stay in the trade.” The opposite holds for a short, which stays open while price closes below the line. Everything else is built around that rule.

• The close counts, not the intraday touch. A bar that dips through the line and closes back above it does not end a long trade under the rule. A second page of the same site, on trading the T-Line, also looks at the next day’s close to confirm a break.
• T-Line runs. A series of closes on the same side of the line is read as a run; a run that ends with a close on the other side is the exit signal.
• The line as support or resistance. In an uptrend, pullbacks to the T-Line that close above it are read as places to add or enter; in a downtrend, rallies that close below it are read the other way.
• Time frames. The author writes that the T-Line works best on daily, weekly and monthly charts and “is not as reliable on the 1- or 5-minute charts.”

The T-Line is often paired with a slower average, such as the 20 or 50-day, to judge the larger trend, which is the idea behind a moving average crossover. On its own it says nothing about where to put a profit target.

Example on the S&P 500 ETF in 2026

The chart above shows the last 120 daily bars of the S&P 500 ETF, from May 8 to October 28, 2026, with the 8 EMA. Price closed above the T-Line on 72 of those 120 days and crossed it 24 times, about once a week on average. The longest run above the line lasted 13 sessions, from June 29; the longest run below it lasted 7 sessions, from August 18. On October 28 the ETF closed at 772.31, just above a T-Line of 768.48.

That is the T-Line in practice: in a steady advance, price spends most days above it, and the 24 crosses are the cost of a line that reacts fast.

Does the T-Line rule work? A test over the full history

We ran the long-only version of the rule on two daily series, over all the history we hold for each, starting flat once the averages were complete. Each day, if the close is above the EMA, the position is long from the next open; if it is below, the position is flat from the next open. Cash earns nothing, there are no costs in the first columns, and the same rule is shown with a 20-day EMA for comparison. Prices are adjusted for dividends and splits, so buy and hold includes dividends.

Three things stand out.

• On the S&P 500 ETF, the rule lost to holding. It was long 62.5% of the days and still lost most of the market’s return: the median trade lost 0.28%, and the many small losses from whipsaws ate the gains of the long runs. Split in two halves at January 2010, the rule made −0.8% a year in the first half against 7.7% for buy and hold, and 4.7% in the second against 14.3%.
• On Bitcoin, the rule beat holding before costs. It was long 52.4% of the days, avoided much of the deepest falls, and kept the big runs. The gap came mostly from the first half: split in two halves at April 24, 2022, the rule made 78.7% a year against 53.1% before, then 18.3% against 17.3% after; the second-half gap depends on the exact split day.
• Costs matter. With an assumed cost of 0.05% per entry or exit on the ETF, the T-Line rule falls to −0.7% a year; with 0.1% per side on Bitcoin, it falls to 35.0%, about level with buy and hold. These cost levels are assumptions for the illustration, not quotes from a broker.

This is one rule tested on two markets, without taxes, slippage or interest on cash, and the Bitcoin result rests on a single long boom. It is also sensitive to the start date: starting the Bitcoin test on other dates between September 6 and November 26, 2017 moves the rule's yearly return between 41.7% and 51.0% and buy and hold's between 28.4% and 42.6%, and with 0.1% costs per side the rule ends behind buy and hold for every start from September 6 to September 25, 2017 (for example 37.5% against 42.6% from September 15). It shows how the T-Line behaves, not that it gives an edge. Our bench methodology explains how we test chart patterns with the same care.

T-Line versus other moving averages

• T-Line versus the 20 EMA. The slower line crossed less often, so it made fewer trades: 542 against 869 on the ETF. It did better there, and about the same on Bitcoin.

On the Bitcoin chart, the 8 EMA turns several bars before the 20 EMA at each swing: that early turn is what gets the T-Line out sooner, and also what makes it whipsaw more in a flat market.

• T-Line versus a simple 8-day average. A simple moving average weighs the 8 closes equally, so it reacts a little later to the newest close.
• T-Line versus a trendline. A trendline is drawn by hand through swing lows or highs; the T-Line is computed and leaves no choice to the chartist.

More trend tools are listed in trend indicators.